furyhawk/llama.cpp
1
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2026-07-21 10:15:53 +00:00
Code Issues Packages Projects Releases Wiki Activity

Fix crash with draft-simple (#25720)

Browse Source 
* Fix crash with draft-simple

* Fix tests for spec decoding
This commit is contained in:
Gaurav Garg
2026-07-15 19:51:34 +05:30
committed by GitHub
parent a582222290
commit 956973c764
3 changed files with 14 additions and 5 deletions
Download Patch File Download Diff File Expand all files Collapse all files
common/speculative.cpp
+4 -1
View File
@@ -260,7 +260,10 @@ struct common_speculative_impl_draft_simple : public common_speculative_impl {
bool process(const llama_batch & batch) override {
auto * ctx_dft = params.ctx_dft;
const int ret = llama_decode(ctx_dft, batch);
llama_batch batch_dft = batch;
batch_dft.logits = nullptr;
const int ret = llama_decode(ctx_dft, batch_dft);
if (ret != 0) {
SPC_ERR("failed to decode draft batch, ret = %d\n", ret);